Rating F1 Drivers

[devilsinskirtsisback 0]

Rightydokey folks, I know that several on here have tried their own/posted others methods of rating drivers. I have to say that I have never seen a method that entirely convinced me that it was spot on. Well, as a statto/analyst type chappie, I have been having a bash at my own figures.

Have I solved the problem (if indeed it is a problem!)?

No.

But, I am intrigued by this first bash at it. Now, before I tell you the process I have used, I would like to post the figures which I have arrived at. I examined all races in the past 7 seasons, the higher a number is the better a driver performed and that's all I'm going to tell you at the moment. I've only included figures here for drivers with a few seasons under their belts. Plus a couple of retired drivers for fun.

Driver 2007 2006 2005 2004 2003 2002 2001

Alonso 22.8 37.8 33.6 14.8 14 2

Barrichello 2 -2 -1.4 13.8 10 11.4 1.2

Button -0.8 19.2 16.4 24 6.4 0.8 -6.6

Coulthard 0.8 1.8 9.8 0.8 3.2 8.2 19

Fisichella 1.2 3.4 4.6 12.4 3.4 4.4 8.6

Heidfeld 14.2 6.6 5.6 4.6 2.2 2.4 3.4

M Schumacher 31.2 21.4 38.6 21.6 32.8 34.6

Massa 13.8 9 8.2 -5.6

Montoya 4.2 14 15.6 17.4 14 8.2

R Schumacher2 6 7 4.8 11.6 4.4 7.8

Raikkonen 27 17 21.4 13 25.2 4.8 0.8

Sato 5.8 4 -8.8 -0.2 -0.4 -2.6

Trulli 0.6 1 10.6 10.2 3.6 3.8 3.4

Villeneuve -1.6 -4.2 -2 -0.8 0.8 5.4

Webber 4 5.4 8.2 8.4 8.4 4.4

So, a first stab, with a few points to like and a few points to hate. Your thoughts?

Let's try that again.!!!

[devilsinskirtsisback 0]

Nope still looks Sh#te. Look for the spaces people.

[rumblestrip 0]

Lemme give it a try...

Driver	   2007   2006   2005   2004   2003   2002   2001
Alonso	   22.8   37.8   33.6   14.8	 14	  2
Barrichello	 2	 -2  -1.4	13.8	 10   11.4	1.2
Button	   -0.8   19.2   16.4	 24	6.4	0.8   -6.6
Coulthard	 0.8	1.8	9.8	0.8	3.2	8.2	 19
Fisichella	1.2	3.4	4.6   12.4	3.4	4.4	8.6
Heidfeld	 14.2	6.6	5.6	4.6	2.2	2.4	3.4
M Schumacher		31.2   21.4   38.6   21.6   32.8   34.6
Massa		13.8	  9	8.2   -5.6
Montoya			  4.2	 14   15.6   17.4	 14	8.2
R Schumacher	2	  6	  7	4.8   11.6	4.4	7.8
Raikkonen	  27	 17   21.4	 13   25.2	4.8	0.8
Sato		  5.8	  4   -8.8   -0.2   -0.4   -2.6
Trulli		0.6	  1   10.6   10.2	3.6	3.8	3.4
Villeneuve		  -1.6   -4.2	 -2   -0.8	0.8	5.4
Webber		  4	5.4	8.2	8.4	8.4	4.4

How'd I do?

[rumblestrip 0]

Heh, I tried to add alternating row colors, but it wasn't too keen on that.

[Grabthaw the Hammerslayer 4]

Er not sure what that proves really - as removing drivers who are short-term changes the basis of the calculation, surely??? e.g. Hammy's not in there so the figures for 2007 are bound to be skewed...

As Vic Reeves once said; 84.7% of all statistics are made up

[Rainmaster 7]

By my calculations Sato is better than Yamamoto, Trulli is better than Sato, Fisi is better than Trulli, Button is better than Fisi, Kimi is better than Button, Hamilton is better than sliced bread, Alonso is better at Renault, and all of them are much better than me.

[dribbler 6]

As Vic Reeves once said; 84.7% of all statistics are made up

Yes but 92.3% of the time, that calculation is wrong.

[Quiet One 15]

That was 68.5% funny!

[Elizabeth Sterling 0]

I have this great method of rating a driver and team's combined performance over the year called the driver's world championship. You see for every top 8 finish a driver scores points, more for the better the finish. First gets 10, second gets 8... and if I tell you anything else it'd just be giving it away which I don't want to do right now.

[Max Mosley 2]

Well, Devilinskirts, I don't know what you want us to say at this stage. Choosing a method based on whether it agrees with our intuition seems like a biased procedure to me.

I have this great method of rating a driver and team's combined performance over the year called the driver's world championship. You see for every top 8 finish a driver scores points, more for the better the finish. First gets 10, second gets 8... and if I tell you anything else it'd just be giving it away which I don't want to do right now.

That's a rubbish method. And it's been tried before too.

[Elizabeth Sterling 0]

Really? Well, crap... okay, how about we reward outright speed by giving 9 for a win, 6 for a second and so forth but only taking the best 11 results? I doubt anyone's done that yet...

[Max Mosley 2]

We should all just trust my intuition and judgement. It's what I do.

[Medilloni 6]

I have a great method too.

I watch them like I'm in the car with them.

On my logging system I record how my hairs (not pubic, that's for something else) stand on end - matched to the ratio of goose bumps (mine, not theirs).

I then work out how many times I've watched my favourites with my fingers crossed for nearly two hours, multiplied that by the degree of cramp in my fingers, then cross reference the bollockings I've given people for walking in/talking/interupting (apart from food/beverage trips etc) whilst the race is on.

Finally, I watch their attitude during interviews etc. I then place them into sub categories: 'Whinging mard little farts', 'shut the fook up and they talk on the track', 'likeable and able to give a good image of the 'sport' to the knobs at work that say it isn't' and 'mean mother fcukers that are ruthless'.

Then, my top drivers are allocated points for their skill at the wheel, how they interact with their team, how much effort they put in between the races at the factory etc.

Finally, if my wife fancies a particular driver, or worse, thinks they're 'cute' I then deduct points for said driver.

Then I................

I'm not telling anymore. No-one else is.

I have four drivers that I really like to watch. The rest make the field up.

[Medilloni 6]

We should all just trust my intuition and judgement. It's what I do.

Priceless. I love that Murray. It's the only way to make any kind of sense!

F1 definately falls into the category of Sex, Religion and comments on your dearest one's driving - it is completely useless to try and reason. It's just what floats people's boat.

[dribbler 6]

That was 68.5% funny!

I love you 35%

I have a great method too.

I watch them like I'm in the car with them.

On my logging system I record how my hairs (not pubic, that's for something else) stand on end - matched to the ratio of goose bumps (mine, not theirs).

I then work out how many times I've watched my favourites with my fingers crossed for nearly two hours, multiplied that by the degree of cramp in my fingers, then cross reference the bollockings I've given people for walking in/talking/interupting (apart from food/beverage trips etc) whilst the race is on.

Finally, I watch their attitude during interviews etc. I then place them into sub categories: 'Whinging mard little farts', 'shut the fook up and they talk on the track', 'likeable and able to give a good image of the 'sport' to the knobs at work that say it isn't' and 'mean mother fcukers that are ruthless'.

Then, my top drivers are allocated points for their skill at the wheel, how they interact with their team, how much effort they put in between the races at the factory etc.

Finally, if my wife fancies a particular driver, or worse, thinks they're 'cute' I then deduct points for said driver.

Then I................

I'm not telling anymore. No-one else is.

I have four drivers that I really like to watch. The rest make the field up.

[devilsinskirtsisback 0]

How'd I do?

You did very well thank you.

To answer some of the points raised:

Meanoni - the calculation is against all drivers who raced that season, hamilton scored 20.8 in 2007. And yes, this would fall into the category of a made up statistic.

Rainmaster - according to my figures Trulli can be better than Fisi.

Elizabeth Sterling - congratulations on your linearity. Does that mean that Alonso was a Sh#te driver until he wasn't a Sh#te driver?

Murray Walker - I was conscious of that possibility before I began and took deliberate steps to avoid it by "agreeing" the formula with myself before I knew any of the results. Indeed, many of the figures I arrived at are opposed to my own intuition.

Medilloni - nice method. I like that. Seriously speaking, this type of thing is only ever going to appeal to those who like this sort of thing. Those who measure drivers on a more intuitive level will never agree with it.

And the method used? The omission of which has clearly caused a degree of upset amongst some.

Simple - compare the record of each driver aginst their own team-mate in races which they both completed. Divide their total points for the season by 5. Add the two figures together.

The reasoning behind the method - wankp**s about favouritism aside, comparison with the team mate is the purest method of assessing the strength of a driver. But it only allows us to measure that person against another at that particular point in time. So whilst Fisi p**sed all over Massa at Sauber, it is unlikely that the same thing would happen if they were in the same team now. Therefore, derived "form" lines cannot be used.

Should points count? Of course they should - the team mates who are level on head to heads but one driver has scored twice as many points, should be reflected in some way.

What do I not like about it? There is no means of rating the achievement of beating a team mate - Sato scores the same for beating Yamamoto 10 times as Kovaleinen would get for beating Fisichella 10 times. Arguably the latter is a greater achievement. And now we are starting to get a little subjective. Also, the same score is achieved for beating your team mate by 1 place, as for beating them by 20 places.

Points form too much of the score.

However, as a starting point, it is reasonable enough.

Finally, does it tell us anything we didn't already know or suspect? Probably not. But **** it. I like it.

[Medilloni 6]

You did very well thank you.

To answer some of the points raised:

Meanoni - the calculation is against all drivers who raced that season, hamilton scored 20.8 in 2007. And yes, this would fall into the category of a made up statistic.

Rainmaster - according to my figures Trulli can be better than Fisi.

Elizabeth Sterling - congratulations on your linearity. Does that mean that Alonso was a Sh#te driver until he wasn't a Sh#te driver?

Murray Walker - I was conscious of that possibility before I began and took deliberate steps to avoid it by "agreeing" the formula with myself before I knew any of the results. Indeed, many of the figures I arrived at are opposed to my own intuition.

Medilloni - nice method. I like that. Seriously speaking, this type of thing is only ever going to appeal to those who like this sort of thing. Those who measure drivers on a more intuitive level will never agree with it.

And the method used? The omission of which has clearly caused a degree of upset amongst some.

Simple - compare the record of each driver aginst their own team-mate in races which they both completed. Divide their total points for the season by 5. Add the two figures together.

The reasoning behind the method - wankp**s about favouritism aside, comparison with the team mate is the purest method of assessing the strength of a driver. But it only allows us to measure that person against another at that particular point in time. So whilst Fisi p**sed all over Massa at Sauber, it is unlikely that the same thing would happen if they were in the same team now. Therefore, derived "form" lines cannot be used.

Should points count? Of course they should - the team mates who are level on head to heads but one driver has scored twice as many points, should be reflected in some way.

What do I not like about it? There is no means of rating the achievement of beating a team mate - Sato scores the same for beating Yamamoto 10 times as Kovaleinen would get for beating Fisichella 10 times. Arguably the latter is a greater achievement. And now we are starting to get a little subjective. Also, the same score is achieved for beating your team mate by 1 place, as for beating them by 20 places.

Points form too much of the score.

However, as a starting point, it is reasonable enough.

Finally, does it tell us anything we didn't already know or suspect? Probably not. But **** it. I like it.

Well Devil, I think you made a damn fine attempt.

I wonder though, what about being as simplistic as (bearing in mind the bit you missed) measuring the time difference between team mates at each race i.e. how many seconds they beat them by.??

Or, the total amount of time it took one team mate against another to cover the entire season - with a complicated formula for DNF's and which of the two got the fastest lap.

[BradSpeedMan 6]

I have a great method too.

I watch them like I'm in the car with them.

On my logging system I record how my hairs (not pubic, that's for something else) stand on end - matched to the ratio of goose bumps (mine, not theirs).

I then work out how many times I've watched my favourites with my fingers crossed for nearly two hours, multiplied that by the degree of cramp in my fingers, then cross reference the bollockings I've given people for walking in/talking/interupting (apart from food/beverage trips etc) whilst the race is on.

Finally, I watch their attitude during interviews etc. I then place them into sub categories: 'Whinging mard little farts', 'shut the fook up and they talk on the track', 'likeable and able to give a good image of the 'sport' to the knobs at work that say it isn't' and 'mean mother fcukers that are ruthless'.

Then, my top drivers are allocated points for their skill at the wheel, how they interact with their team, how much effort they put in between the races at the factory etc.

Finally, if my wife fancies a particular driver, or worse, thinks they're 'cute' I then deduct points for said driver.

Then I................

I'm not telling anymore. No-one else is.

I have four drivers that I really like to watch. The rest make the field up.

[devilsinskirtsisback 0]

Or, the total amount of time it took one team mate against another to cover the entire season - with a complicated formula for DNF's and which of the two got the fastest lap.

Hmm, I like this idea. There is possibly potential to introduce a factor for the car's input to the times as well. Hmmm. I shall ponder further.

[shampion 0]

No Brainer in the results you have..... Driver performance? Maybe but doesn't say to me who is a better driver.

If a car is physically slower than another or doesn't have reliability.... obviously his performance is less. Means nothing about the driver... its more about the car.

As I always said.... the only way drivers can be compared fairly is by giving them all the same machinery let them go at it for several races. In the end.... May the best pilot win!

[Max Mosley 2]

It's a good starting point Devil! Like Med says above, maybe you could look at the times as well or instead of points differences. Also can you weight the points/time difference between team mates by the score of the team mate in some way? Maybe you should look at the ELO rating formula they use in chess rankings? I think the number of points gained by winning a match depends on how highly ranked the opponent was at the start of the match.

[Max Mosley 2]

always said.... the only way drivers can be compared fairly is by giving them all the same machinery let them go at it for several races. In the end.... May the best pilot win!

Yep. That is the way F1 should be imho.

[Medilloni 6]

Yep. That is the way F1 should be imho.

True. True.

Soooooooooooooooooo, what if Devil figures some 'not as good as the rest' formula - a 'handicap' for each car? For example, let's say the your benchmark is the WCC (arguable? On this forum,yes ), then McLaren might be 3% slower, Toro Roso 15%, Honda 12% etc etc. The theory being that it would be as close as you could get for every driver in the same car.

Is that close to your Chess idea Murray? (I only play tiddlwinks)

I think what Devil has done has the potential for great fun. And potential for a true, proper, full blown TF1 'It's Not Fair Argument'

Anyway, I stick by my original post - and I know who's best 'cos the world is wrong 'cept me

[pabloh20 1]

True. True.

Soooooooooooooooooo, what if Devil figures some 'not as good as the rest' formula - a 'handicap' for each car? For example, let's say the your benchmark is the WCC (arguable? On this forum,yes ), then McLaren might be 3% slower, Toro Roso 15%, Honda 12% etc etc. The theory being that it would be as close as you could get for every driver in the same car.

Is that close to your Chess idea Murray? (I only play tiddlwinks)

I think what Devil has done has the potential for great fun. And potential for a true, proper, full blown TF1 'It's Not Fair Argument'

Anyway, I stick by my original post - and I know who's best 'cos the world is wrong 'cept me

I am hesitant to disagree with my esteemed colleague, but if you excuse the interludity, I feel I must interject the correct resolution of the boldifacated statement above. I have one small amendment to make, if you pardon the contrifibularity.

Anyway, I stick by my original post - and I know who's best 'cos the world is wrong 'cept my wife

There, that's better!

[shampion 0]

Anyway, I stick by my original post - and I know who's best 'cos the world is wrong 'cept me

I agree with you!... but if I am wrong then you are wrong...